Medicare's End-Stage Renal Disease (ESRD) program has enjoyed immense success, supporting the lives of over 500,000 Americans since its inception in 1973 and currently maintaining approximately 210,000 (1993 census), most by hemodialysis. Despite this success, concern has been raised about possible excess mortality in patients enrolled in the hemodialysis program. These concerns are fueled by reports from several other large countries where mortality is lower, although the patient mix is usually different. Higher mortality rates in the U.S have been attributed to more liberal patient eligibility as evidenced by higher incidence rates, but suspicion has been raised that inadequate dialysis may play a role. The last major controlled study of hemodialysis in the U.S., the National Cooperative Dialysis Study (NCDS), 1976-1980, showed unequivocally that morbidity and mortality could be reduced by more intensive dialysis that targeted removal of urea. The NCDS re-focused attention on removal of small molecules like urea but it was not designed to examine the more intensive solute removal now possible with modern dialyzers. In addition, the study excluded several major categories of patients currently receiving dialysis treatments (e.g., diabetics) and it examined none of the high-flux, more biocompatible dialyzers now widely used. The present study proposes to determine the effects of increasing the dialysis dose from a currently acceptable level, measured as whole body Kt/V=0.95/dialysis, to a level of 1.4/dialysis in 900 patients from 15 dialysis centers dialyzed thrice weekly over a period of seven years. Kt/V is a measure of the dose of dialysis best defined as the fractional volume dialyzed per treatment, measured with a two-pool model of urea kinetics. The -50% increment in dosage will be combined, using a 2 x 2 factorial design, with a comparison of cellulose-derived dialysis membranes and more porous, biocompatible dialysis membranes. The goal is to compare mortality among the four groups and to compare morbidity using end-points such as number and frequency of hospitalizations, nutritional parameters, and subjective health status assessment. Eligibility for enrollment in this study will be less strict than the previous study but will include far advanced and well-established end-stage renal failure, age equal to or greater than 18 and equal to or greater than 75 years, absence of severe immediately life-threatening diseases, and the willingness and ability to follow a medical regimen. These studies will answer important questions about optimal quantity and quality of hemodialysis to improve the average health and life expectancy of patients with ESRD as well as optimize the economic impact of treatment. Additional health and economic benefits may result from reduction in hospitalizations, fewer diagnostic procedures, and less need for vascular surgery if more intense dialysis and/or biocompatible membranes have significant benefits.